FNZ 68 - Simuliidae (Insecta: Diptera) - Geographic distribution and altitudinal range
Craig DA, Craig REG, Crosby TK 2012. Simuliidae (Insecta: Diptera). Fauna of New Zealand 68, 336 pages.
(
ISSN 0111-5383 (print),
ISSN 1179-7193 (online)
;
no.
68.
ISBN 978-0-478-34734-0 (print),
ISBN 978-0-478-34735-7 (online)
).
Published 29 June 2012
ZooBank: http://zoobank.org/References/9C478D54-FEB2-45E8-B61C-A3A06D4EB45D
Geographic distribution and altitudinal range
New Zealand Austrosimulium species occur on the three main islands (North, South, and Stewart) and also on Great Island of Three Kings Islands, Great Barrier Island, and the subantarctic Auckland Islands and Campbell Island. They are not present on the Chatham Islands (Dumbleton 1973). West of Stewart Island they are known to occur on Big South Cape Island and Codfish Island, an indication that simuliids probably occur on other small islands too.
The size of island that can support a simuliid fauna is of some interest in biogeography of Simuliidae, because they are known from a variety of isolated islands world-wide (Adler & Crosskey 2012). Minimum size of island was discussed by Craig (2003, his Fig. 2) for Polynesian islands of hot spot origin, and involved the interaction of age, altitude, and size of an island in relation to the given climate and the ability to produce running water. If rainfall is sufficient, very small islands can support simuliids (Craig 2004).
A biting adult female A. australense was collected on Little Barrier Island by C. T. Duval on 6 April 1984, but he did not find any immature stages (NZAC record). Neither did Wise (1956) nor Winterbourn (1964). Similarly TKC found none in February 1976 when streams had very low to no water flow. But K. M. Crosby reported to TKC that while on Little Barrier Island in January 2009 he was bitten by simuliids, but did not, however, collect any specimens. These two records of biting relate to comments below on dispersal of aquatic insects (p. 42, Bunn & Hughes 1997).
The minimum size for New Zealand islands to support simuliids is not known, but we suggest that Great Island of the Three Kings is close to that for a species in which larvae are able to cope with very low water flow, and the much smaller Cuvier Island on which Riddell (1981) failed to find simuliids is below the limit. Little Barrier Island, would appear to be close to the limit for a species requiring constant water flow characteristic of most simuliid larvae.
Distributions of New Zealand Austrosimulium species (Maps 1–19, and summarised in Table 1) are plotted over a system of aquatic ecological regions for New Zealand (Map 20) as proposed by Harding & Winterbourn (1997a). Their ecoregions were based on overlapping parameters of geology, altitude, rainfall, and vegetation based on criteria of Biggs et al. (1990). They refined these regions for the South Island (Harding & Winterbourn 1997b) and then (Harding et al. 1997b) examined the relevance of their ecoregions to aquatic faunal assemblages, but only for the South Island. These ecoregions have been used in studies such as Duggan et al. (2002) and Winterbourn (2008). Where possible we discuss distribution of Austrosimulium spp. to these ecoregions. Although temperature is embodied in these ecoregions, we sometimes comment specifically on distribution in relation to mean annual temperatures in New Zealand (NIWA 2008).
The 2-letter codes for the aquatic ecological regions proposed by Harding & Winterbourn (1997a, b) and used on Maps 1–20 and Table 1 are as follows:
North Island: BP, Bay of Plenty Lowlands; CL, Coromandel Peninsula; CM, Central Mountains; EH, East Cape Highlands; EL, Eastern Arable Lowlands; HP, Hauraki Plains; MN, Manawatu Plains; ND, Northern Hill Country; TK, Mt Taranaki Forest; TO, Taupo Plateau; VP, Volcanic Plateau; WA, Wairarapa Highlands; WO, Waikato Hill Country. South Island: CO, Central Otago; EC, East Coast Plains; HC, High Country; MP, Marlborough Plains; NE, Northeast Nelson Forest; NP, Nelson Plains; NN Northwest Nelson Forest; PE, Banks Peninsula; SA, Southern Alps; SE, Southeast Forest; SL Southland Plains; WD, Westland Forest.
Arbitrary area codes proposed by Crosby et al. (1976) to facilitate categorising terrestrial specimen localities, and shown at the end of the monograph, are also provided in Appendix 2 and as a summary in Table 2. These area codes were updated by Crosby et al. (1998). They were not intended to be biogeographical regions although they followed broad climatic regions used in weather forecasts. Files containing these area codes as a data layer for use in GIS programs and Google Earth are available without cost from <http://lris.scinfo.org.nz/layer/165-nz-area-codes-for-recording-specimen-localities/>.
The distribution of the 19 New Zealand species of Austrosimulium recognised is disparate. While species in the australense species-group are found on both islands, only 4 are known from the North Island, 3 of those in common with the South Island, and the fourth, A. dugdalei, precinctive to central North Island. Austrosimulium dugdalei is, however, closely related to A. multicorne of the South Island. All species in the ungulatum species-group are precinctive to the South, Stewart, and the subantarctic Auckland and Campbell Islands. The distribution of species by ecoregion (Table 1), as with the ecological overlay on the reconstructed phylogeny (Fig. 507), well shows that some species segregates are associated with cold water and high altitude.
australense species-group
australense-subgroup
The most widely distributed simuliid is A. australense. It is found, literally, from the most northeastern part of Northland to the southernmost edge of South Island and onto Stewart Island (Map 3). It is particularly widely spread in the North Island. Absences are in the Wairarapa Highlands aquatic ecoregion, that of the Hauraki Plains, and the central part of Waikato Hill Country. The first two absences appear related to lack of suitable running water, in large part due to intense agriculture, whereas the latter is related to lack of collecting. Recent collections indicate that A. australense does occur in Waikato Hill Country, however sporadically and at low frequency. Similarly, in the South Island there are distributional gaps and unexamined areas.
Larvae of A. australense typically use trailing vegetation as a substrate. Therefore, larger braided rivers that have little trailing vegetation tend not to have this species: this applies to running water habitats of North, Mid and South Canterbury, and Otago areas. An outlier locality in Canterbury was the North Branch, Ashburton River where larvae and pupae were recovered from trailing branches of willow trees; we consider this record provides an indication that A. australense may occur elsewhere in that area, but with spotty distribution related to riparian vegetation. Sinton (2008) reported simuliids from irrigation (stock) water races on the Canterbury Plains, but did not identify the species. In general A. australense occurs sporadically along the ecoregions of Westland Plains, north in Nelson Plains, and Northeast Nelson Forest. It is, with few exceptions, absent from those of Marlborough Plains, High Country, Eastern Plains, Central Otago, and Southern Alps; however, it is arrayed along the southern edge of the South Island and positively identified from a few localities on Stewart Island.
Dumbleton noted that A. australense was restricted to an upper altitude of 610 m a s l on the Volcanic Plateau, North Island. This does seem to be the general upper limit, since we recovered A. australense adults from 680 m a s l on Mount Taranaki; however, we have collected immatures at 900 m a s l on Mount Ruapehu, Ohakune. While we have collected this species from near sea level, the majority of localities lie between 100–200 m a s l (Appendix 1).
Molecular analysis indicates that A. australense is probably two separate cryptic species. One clade is exclusively in the South Island. The other clade is almost completely in the North Island, but with apparent exemplars in the South Island; perhaps indicative of precinctive evolution of that clade in the North Island with subsequent dispersal to the South Island (see p. 63 and 77, Molecular Analysis and Biogeography sections).
The related species A. longicorne is also found on both main islands and similarly the molecular lineages differ on each island — a single one in the North Island. This species has specialised habitat requirements and, although widespread, is not common (Map 11). However, recent collections indicate that the species is more common than previously thought. Austrosimulium longicorne is the northernmost species found, occurring on Great Island of the Three Kings Islands (not the common A. australense as might be assumed). However, while A. longicorne seems to require specific habitats (see species description), it can survive in intermittently-flowing streams (e.g., Fig. 472, 473), and these occur on the Three Kings Islands.
tillyardianum-subgroup
The most widely distributed species in this subgroup is A. tillyardianum (Map 14), but it is restricted to south of Auckland in the North Island and north of Dunedin in the South Island. The northern distribution may well be determined by temperature, falling almost exactly within small areas of 12–14°C mean annual temperature (NIWA 2008). However, annual temperatures in the 14–16°C range do occur around the coast of East Cape where A. till-yardianum is found. There is a gap in the mid Waikato Hill Country aquatic ecoregion, which relates to both a paucity of collecting and the substrates in streams and rivers there tending to be of soft papa (blue mudstone); not suitable for A. tillyardianum larvae, which typically are on hard substrates. The absence of A. tillyardianum from Mount Taranaki Forest aquatic ecoregion, and the Manawatu Plains, is something of a puzzle, but rivers arising on the mountain are torrential and not particularly suitable. Lower reaches, even on the Manawatu Plains, are compromised by severe and intermittent flooding and unstable substrates that again are somewhat unsuitable. Why it is A. australense that occurs on this mountain and at high altitude is a further puzzle. In the South Island the distribution and cut-off north of Dunedin of A. tillyardianum may also represent a temperature limit, with most populations occurring in areas of 10–12°C mean annual temperature. That agrees in part with distribution of A. tillyardianum on the Volcanic Plateau, North Island, which mainly falls into a similar temperature regime although at higher altitude, but lower latitude. On the east coast of the South Island, the majority of localities occur in the East Coast Plain, Banks Peninsula, and High Country ecoregions. Localities farther north are mainly in the Northeast Nelson Forest and Nelson Plains.
Both Tonnoir (1925) and Dumbleton (1973) considered A. multicorne to occur on both islands. We have, however, raised the North Island segregate of A. multicorne to specific status, namely A. dugdalei Craig, Craig & Crosby. There is no question that the two species are closely related, with A. dugdalei recently derived from A. multicorne. We make the assertion that since A. dugdalei occurs only on the Volcanic Plateau (Map 6), and this ecoregion is no more than 1.6–2 million years old (Thornton 2003; Campbell & Hutching 2007; Graham 2008), then divergence of these taxa has occurred within this period of time. Known from only 11 localities, the majority of collections are on the Volcanic and Taupo Plateaus (VP, TP), with one farther south at the northern edge of the Manawatu Plains. Austrosimulium dugdalei is restricted to an altitude distribution of 500–1 150 m a s l. The localities fall well within mean annual temperatures of 6–12°C. The species might also occur farther northeast on the Taupo Plateau, such as in the Kaimanawa Mountains, but access is difficult.
Austrosimulium multicorne in the South Island is well distributed and occurs in all areas (Map 12), and there are problems with discrimination between it and A. stewartense in the southern part of the island (see species description). The majority of localities fall into the High Country ecoregion and the species, with few exceptions, is absent from East Coast Plains. A surprise is that A. multicorne occurs on Banks Peninsula, where it was not recovered by Crosby (1974a), even during his intensive long term study, probably because A. multicorne occurs at higher altitudes. The ecoregions of Banks Peninsula and the High Country cluster together in terms of climatic and geomorphological characteristics (Harding & Winterbourn 1997b), so ecological parameters are suitable. Dumbleton (1973) gave the upper limit of A. multicorne as 1 200 m a s l at Mount Balloon, NW Nelson. We have collected it at 1 400 m a s l at Temple Basin, Arthurs Pass, and there other records of it from 1 600 m a s l, on the Old Man Range, Otago. While it has been collected close to sea level, it is normally found at 200–600 m a s l (Appendix 1), where spot temperatures range from 8 to 18°C, with most from 8 to 15°C. This is in reasonably good agreement with mean annual temperature of the High Country ecoregion between 6 to 10°C.
Farther south in the South Island is the poorly known A. fiordense, originally considered by Dumbleton to be a subspecies of A. multicorne, and raised here to full species status. Beyond the type material from the Glasinock and Stillwater Rivers, Lake Te Anau, it is known also from high altitude in the Darran Mountains, Fiordland (both in the Southern Alps ecoregion) and an isolated population from near Mesopotamia, Rangitata River Valley (High Country ecoregion) where it occurred with A. multicorne and A. tillyardianum (Appendix 1, Map 9).
Austrosimulium stewartense has a southern distribution (Map 13). The localities in the areas of the Mid Canterbury foothills and farther south in Central Otago (High Country ecoregions) are moot. As known, this species is, at some locations, hard to separate from A. multicorne. However, all localities fall into ecoregions with mean annual temperatures of 8–10°C. Those in the south are mainly in the Southland Plains, but they are also in the Southeast Forest and Westland Plains. Dumbleton gave no altitudinal limits, but his label data indicate 20–400 m a s l, and we have collected it from literally high tide level up to some 760 m a s l (Appendix 1). On Stewart Island, Chadderton (1990) collected aquatic invertebrates extensively around the northern half of the island and sporadically farther south. Austrosimulium was amongst the 5 most abundant taxa and occurred in 41 of the 45 sites samples. The species were not identified, but we assume most were A. stewartense.
Now separated from A. stewartense as a distinct species, A. extendorum Craig, Craig & Crosby has a restricted distribution, known from only 2 localities, one on Stewart Island, another on nearby Big South Cape Island (Map 8). Austrosimulium extendorum, along with A. stewartense, is unique; larvae have been found on cobbles at high tide mark (Fig. 466).
Austrosimulium laticorne is precinctive to the South Island (Map 10) and has a distribution not uncommon to South Island species. Localities are concentrated in the ecoregions of Northeast Nelson Forest, Northwest Nelson Forest, and Nelson Plains. Older collections show it to be arrayed centrally along the Westland Forest ecoregion. A single questionable locality is in the Mackenzie High Country. There is then a distinct gap until south of Dunedin, where it is quite common in the ecoregions of Southland Plains and along the edge of the High Country. Overall, the distribution indicates a preference for a mean annual temperature of 8–12°C. Recorded spot temperatures (Appendix 1) are, however, warmer and range from 10 to 23°C. The major gap in distribution, north of Dunedin and up to the latitude of Timaru, indicates that there might be two taxonomic entities, reminiscent of the distribution of A. australense in the South Island. Dumbleton made no comment about altitudinal distribution, but we have collected A. laticorne from sea level to 826 m a s l (Appendix 1).
The closely related A. alveolatum is of limited distribution, occurring only in the Mid Canterbury area (Map 2) corresponding to the central portion of the High Country ecoregion. There is a narrow altitudinal range from 250–706 m a s l in our collections, but Dumbleton’s label data indicated an upper range of 920 m a s l. The localities tend to fall into the mean annual temperature range of 6–10°C, but the spot temperatures recorded were of 11–15°C (Appendix 1).
Austrosimulium albovelatum has a similar distribution in many ways to A. alveolatum, but with two exceptions. One known population is in the Mid Canterbury area (Map 1), the other close to sea level, at Kaikoura (Appendix 1). Of importance here is that all localities fall into the High Country aquatic ecoregion and that includes Kaikoura. Along the Seaward Kaikoura Mountains, cold-water alpine streams empty directly into the ocean. The altitudinal range of A. albovelatum is normally narrow, 250–616 m a s l. The temperature regime for the High Country ecoregion is 6–10°C, but spot temperatures for localities were higher, 12–18°C (Appendix 1).
ungulatum species-group
No members of the ungulatum species-group are found in the North Island. The reasons for this have been the subject of considerable speculation (Tonnoir 1925; Dumbleton 1973). Dumbleton specifically noted that the group is precinctive to South and Stewart Islands, and the subantarctic islands. The absence from North Island is surprising given the ability of A. ungulatum to fly kilometers (Tonnoir 1925; Craig pers. obs.). There is a expectation that with sea level depression during glaciations, it would have been a minor dispersal event to cross the narrowed or even closed Cook Strait to reach Mount Taranaki, in particular. There are ample cold-water, densely-shaded streams on that mountain, suitable for A. ungulatum. But no. Some authors (e.g., Worthy & Holdaway 2002) were not convinced that land fully connected the North and South Islands during the last glaciation — Cook Strait was just markedly narrowed. Further, there is evidence that weather systems, and in particular, the “Roaring Forties” were shifted northwards and intensified during glaciations (Newnham et al. 2003; Carter & Gammon 2004), making a crossing of even a reduced Cook Strait more fraught even than at present.
ungulatum-subgroup
Austrosimulium ungulatum is found in all parts of the South Island (Map 16). There is molecular evidence that A. ungulatum is perhaps a complex of at least two species (see p. 64, Molecular Analysis section) and Dumbleton (1973) noted marked variation in gill filament number. In terms of aquatic ecoregions it is absent from the East Coast Plains, with one exception, however, it abuts the western edge of that ecoregion and the eastern High Country. The exception was 2 female adults recovered from Kaituna Bush, Banks Peninsula. Of note is that the typical habitat for A. ungulatum larvae is small, densely-shaded, cold water streams. Larvae occur at low density and Tonnoir (1925) did not discover them; it was Dumbleton (1973) who first found them. In this instance, Kaituna Bush, while physically in the East Coast Plains ecoregion, has Kaituna River flowing through it from the Banks Peninsula ecoregion: the upper reaches of the stream are heavily shaded and cold; the Bush reserve lower down is surrounded by agricultural land. Why was A. ungulatum not collected on Banks Peninsula previously? Was this collection from a normal dispersal event, unsuccessful because of competitive exclusion by A. tillyardianum? It poses a similar problem to that with A. multicorne which also occurs on Banks Peninsula at extremely low frequency (Map 12, Appendix 1). Or are both recent immigrants? We note elsewhere that Banks Peninsula has been heavily impacted by human activity and currently has little of the original forest cover.
The presence of A. ungulatum at the Kaikoura coast might seem unusual, but again the High Country aquatic ecoregion extends out to the coast there, and some streams are classic habitats for that species (Fig. 493). Austrosimulium ungulatum is sparsely distributed in Central Otago and the southeastern High Country ecoregions. Dumbleton gave an altitudinal range of from sea level to 920 m a s l. We found a similar range, but up to 993 m a s l (Porters Pass, NZS134). The absence of A ungulatum from the Southern Alps ecoregion is notable because of the lack of biting by females of this species (Crosby & Craig pers. obs.); a role filled in part by other species.
The ecoregions for A. ungulatum are in the mean annual temperature range of 6–10°C. Spot temperatures were, however, from 9–23°C (Appendix 1). Exceptions to the requirement for cooler temperatures are those of a population in the Marlborough Sounds area (Northeast Nelson Forest ecoregion) and those in the northwest Nelson area (Nelson Plains ecoregion).
A new species, A. vailavoense Craig, Craig & Crosby is restricted to Stewart Island and the extreme south of the South Island (Map 17). The localities on the South Island are within the Southland Plains, and on Stewart Island in the Southeast Forest ecoregion. Altitude ranges from close to sea level to 21 m a s l. The immature stages are as yet unknown, but the localities known are ecologically similar to those required by immatures stages of A. ungulatum; small, cooler (13–15°C), densely-shaded streams (Appendix 1).
In the subantarctic islands, A. vexans is precinctive to the Auckland Islands, and A. campbellense is on Campbell Island (Map 5, 19). Dumbleton (1973) gave a few details for A. vexans, such as its occurrence at altitudes lower than 90 m a s l and in large numbers when the stream bed was cobble. For Campbell Island, Dumbleton (1973) also gave some details, such as mean air temperatures of 9.4° in January and 4.4°C in June and July. TKC has collected A. campbellense larvae at 220 m a s l. Joy & Death (2000) examined stream invertebrate communities on Campbell Island and of 19 streams examined, island wide, 7 had A. campbellense, all at altitudes below 40 m a s l. Most streams were deeply incised, with a mean velocity of 0.6 m/s, but some were recorded as having zero velocity! Temperatures of the water were all in the 8–9°C range.
Austrosimulium dumbletoni is, on morphological grounds, currently placed in the ungulatum species-group (Crosby 1976a). It is known only from the Westland Forest aquatic ecoregion at Jackson Bay and at Knights Point at sea level when attracted to penguins and humans; although it does not bite humans. Immature stages are unknown. With a probable relationship to species in the unicorne-subgroup (p. 64, Molecular Analysis), similar to those species, the immature stages of A. dumbletoni may occur at high altitude.
unicorne-subgroup
These are high altitude simuliids and of relatively restricted distribution. Austrosimulium bicorne is only found in the southern Westland area and northern Fiordland. There is also a small locus at the western extent of North Canterbury. All these localities fall within the Southern Alps, or are at the western edge of the High Country aquatic ecoregions (Map 4). They have a mean annual temperature of 4–8°C. With one exception, spot temperatures agreed closely with that, ranging from 6–9°C. The exception was 15°C, where water flowed down a sunlit cliff into the stream habitat below (Appendix 1, Fig. 491). The altitudinal range is narrow, from 882–1 600 m a s l. The disjunction between the two populations is probably due to a lack of collecting; not easy at that altitude in the Southern Alps.
Austrosimulium unicorne has a similar ecological distribution to A. bicorne. Found only in the North, Mid, and South Canterbury areas, it is sharply restricted to the eastern side of the Southern Alps aquatic ecoregion and the western edge of the High Country (Map 16). With few exceptions it is only found between 700–1 677 m a s l. Lower altitude records of 20–350 m a s l are suspect and no ecological data are available for these records. Temperature data during our collecting (6–9°C) indicates it is a cold water species. Dumbleton noted that in a typical habitat, even at a lower altitude, winter temperatures were 3°C and during summer 6°C.
A. tonnoiri Craig, Craig & Crosby is another member of this unicorne-subgroup. Dumbleton included the specimens he had from Homer Tunnel as a variant of A. bicorne, but with additional specimens we found this entity should be recognised as having species status. Austrosimulium tonnoiri has a restricted distribution in the Southern Alps aquatic ecoregion similar to that of A. bicorne, but does not have mid-island populations (Map 15). It has a restricted altitudinal range, from 660–1 150 m a s l, and occurs too in markedly cold water (6–9°C) (Appendix 1).
We have noticed that repeated collections at some localities do not necessarily return the same species of Austrosimulium, if any at all; or a site visited unsuccessfully many times suddenly produces simuliids. This is of some significance since it means that distributions are not static. The phenomenon has been well examined by others and an example is that by Bunn & Hughes (1997). In a broadly-based study they investigated the drift downstream and movement upstream of aquatic invertebrates, as well as distance flown by adult insects. They also examined the genetic diversity of both a Baetis Leach mayfly and a Tasiagma Neboiss caddisfly. Their general expectations were — if dispersal of an organism is high there should be little genetic differentiation among populations; on the other hand, with restricted dispersal, differentiation between populations would occur through natural selection or random genetic drift (Slatkin 1985). Against expectations, they found that genetic differentiation was not necessarily a function of geographic distance and that genetic variation in a population, from which multiple samples had been taken over time, was significantly different between times. They also calculated that to produce densities of the insects studied, only a few adults would be needed. Their conclusions were that it is the adults that do the dispersing, and dispersal capabilities and stochastic effects of recruitment can be determining factors in observed spatial and temporal variation in community structure in some streams.
A similar study was by Hogg et al. (2002) on population genetic structure of two New Zealand stream insects, namely the megalopteran Archichauliodes diversus (Walker) and ephemeropteran Coloburiscus humeralis (Walker). Expectations were that A. diversus having markedly large and poor flying adults would show greater genetic divergence between populations because of poor dispersal ability. Similarly this was expected for C. humeralis, but there because the adults were assumed to be short lived. The expectations were met for A. diversus which showed moderate genetic divergence between North Island populations and a major divergence between those and the single population sampled from the South Island. Coloburiscus humeralis on the other hand, did not meet expectations and showed extreme lack of genetic divergence. While that could be attributed to major dispersal ability, Hogg et al. (2002) had other suggestions. One was that recurring glaciation and concomitant sea level fluctuation, major volcanism, and habitat fragmentation resulted in reduction of genetic diversity in remaining small populations. Recolonisation from these genetically depauperate founder populations might then explain the current low genetic diversity in C. humeralis; and for that matter other New Zealand fauna.
The overview by Bunn & Hughes accords with our observations. For example, the Mountain Road, Ohakune site (NZN92) was visited 3 times. The first time (March) produced larvae of A. australense and A. dugdalei. The first species was well outside its usual altitudinal range. The next 2 visits were both in November, 2 and 3 years apart respectively, and produced nothing, yet the stream appeared the same. Similarly, near Waiouru (NZN46, November) the first visit produced A. longicorne, but a visit 2 years later (November) produced no simuliids. The stream at the summit of Takaka Hill on the first visit (NZS103, February) produced only 3 immature larvae of A. longicorne, whereas a second visit 3 years later (November) produced numerous larvae of A. ungulatum, 1 final instar of A. multicorne, and numbers of unidentifiable immature larvae. The presence and absence of simuliids on Little Barrier Island, as discussed above, is a similar situation. Overall, we suggest that species composition at any one locality should not necessarily be taken as definitive from our data.